| { | |
| "best_metric": 0.03905326500535011, | |
| "best_model_checkpoint": "./vit-base-beans-demo-v19Set/checkpoint-200", | |
| "epoch": 4.0, | |
| "eval_steps": 100, | |
| "global_step": 260, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 1.670286774635315, | |
| "learning_rate": 0.00019230769230769233, | |
| "loss": 0.7813, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 2.140129327774048, | |
| "learning_rate": 0.00018461538461538463, | |
| "loss": 0.312, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 1.0461418628692627, | |
| "learning_rate": 0.00017692307692307693, | |
| "loss": 0.1888, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 3.5472192764282227, | |
| "learning_rate": 0.00016923076923076923, | |
| "loss": 0.2257, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 2.7260594367980957, | |
| "learning_rate": 0.00016153846153846155, | |
| "loss": 0.2586, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 4.067043781280518, | |
| "learning_rate": 0.00015384615384615385, | |
| "loss": 0.2372, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 0.5244776010513306, | |
| "learning_rate": 0.00014615384615384615, | |
| "loss": 0.1362, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 0.14199833571910858, | |
| "learning_rate": 0.00013846153846153847, | |
| "loss": 0.2058, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 0.16960681974887848, | |
| "learning_rate": 0.00013076923076923077, | |
| "loss": 0.0822, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 1.8150014877319336, | |
| "learning_rate": 0.0001230769230769231, | |
| "loss": 0.0984, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "eval_accuracy": 0.9849624060150376, | |
| "eval_loss": 0.05019935220479965, | |
| "eval_runtime": 1.261, | |
| "eval_samples_per_second": 105.471, | |
| "eval_steps_per_second": 13.481, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 4.9407758712768555, | |
| "learning_rate": 0.00011538461538461538, | |
| "loss": 0.1108, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 0.2821877598762512, | |
| "learning_rate": 0.0001076923076923077, | |
| "loss": 0.1016, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.10290689021348953, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0856, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 0.24183936417102814, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.0728, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 1.1455119848251343, | |
| "learning_rate": 8.461538461538461e-05, | |
| "loss": 0.0437, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 0.09371420741081238, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.028, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.6153846153846154, | |
| "grad_norm": 6.250518321990967, | |
| "learning_rate": 6.923076923076924e-05, | |
| "loss": 0.0765, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "grad_norm": 2.493722438812256, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.0899, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.9230769230769234, | |
| "grad_norm": 1.267454743385315, | |
| "learning_rate": 5.384615384615385e-05, | |
| "loss": 0.0285, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 0.06802109628915787, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 0.0388, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "eval_accuracy": 0.9774436090225563, | |
| "eval_loss": 0.03905326500535011, | |
| "eval_runtime": 2.6673, | |
| "eval_samples_per_second": 49.864, | |
| "eval_steps_per_second": 6.374, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.230769230769231, | |
| "grad_norm": 0.06581737101078033, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 0.0146, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.3846153846153846, | |
| "grad_norm": 0.06294853240251541, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.0138, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.5384615384615383, | |
| "grad_norm": 0.06758769601583481, | |
| "learning_rate": 2.307692307692308e-05, | |
| "loss": 0.0202, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 3.6923076923076925, | |
| "grad_norm": 0.06321461498737335, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.0119, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.06657669693231583, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.0139, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.1454625427722931, | |
| "learning_rate": 0.0, | |
| "loss": 0.0122, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 260, | |
| "total_flos": 3.205097416476426e+17, | |
| "train_loss": 0.1264910431435475, | |
| "train_runtime": 114.5573, | |
| "train_samples_per_second": 36.104, | |
| "train_steps_per_second": 2.27 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 260, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.205097416476426e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |